Breast cancer is a disease for which essentially no practical preventive measures have been established. Recent evidence suggest that both alcohol consumption and lactation are determinants of breast cancer risk, with modest but potentially important magnitudes of effect. Further study of these modifiable exposures would provide more information on which women and their physicians can base decisions regarding behavior, and may provide insights into the biology of the tumor. A large population-based case-control study is proposed to evaluate the relationships between age-specific alcohol consumption and lactation and the occurrence of breast cancer. Specific hypotheses to be addressed include (1) cessation of alcohol intake is associated with a lower risk of breast cancer than is continued consumption, (2) alcohol intake before age 30 is associated more strongly with increased risk of breast cancer than is consumption at later ages, and (3) lactation is associated with a decreased risk of breast cancer among premenopausal women. To test these and other hypotheses, telephone interviews with approximately 5500 women with newly diagnosed breast cancer and 5500 randomly selected female community members will be conducted. In each of the three participating states (Maine, Massachusetts, and Wisconsin) cases will be identified through state tumor registries. Controls will be selected from general population lists (state driver's license lists and Medicare beneficiary lists). A large sample is needed since (1) the associations of interest are modest and (2) women with the alcohol consumption pattern required to test a key hypothesis are infrequent. The study is collaborative. Both Harvard School of Public Health and the University of Wisconsin Clinical Cancer Center have submitted separate applications with identical research plans.